Affordable Access

How might we model an ecosystem?

Authors
Journal
Ecological Modelling
0304-3800
Publisher
Elsevier
Publication Date
Volume
220
Issue
17
Identifiers
DOI: 10.1016/j.ecolmodel.2009.04.021
Keywords
  • Model Prediction
  • Simulation Models
  • Food Web
  • Functional Ecology
  • Trophic–Dynamics
Disciplines
  • Biology
  • Ecology

Abstract

Abstract Predicting ecosystem effects is of crucial importance in a world at threat from natural and human-mediated change. Here we propose an ecologically defensible representation of an ecosystem that facilitates predictive modelling. The representation has its roots in the early trophic and energetic theory of ecosystem dynamics and more recent functional ecology and network theory. Using the arable ecosystem of the UK as an example, we show that the representation allows simplification from the many interacting plant and invertebrate species, typically present in arable fields, to a more tractable number of trophic-functional types. Our compound hypothesis is that “trophic-functional types of plants and invertebrates can be used to explain the structure, diversity and dynamics of arable ecosystems”. The trophic-functional types act as containers for individuals, within an individual-based model, sharing similar trophic behaviour and traits of biomass transformation. Biomass, or energy, flows between the types and this allows the key ecological properties of individual abundance and body mass, at each trophic height, to be followed through simulations. Our preliminary simulation results suggest that the model shows great promise. The simulation output for simple ecosystems, populated with realistic parameter values, is consistent with current laboratory observations and provides exciting indications that it could reproduce field scale phenomena. The model also produces output that links the individual, population and community scales, and may be analysed and tested using community, network (food web) and population dynamic theory. We show that we can include management effects, as perturbations to parameter values, for modelling the effects of change and indicating management responses to change. This model will require robust analysis, testing and validation, and we discuss how we will achieve this in the future.

There are no comments yet on this publication. Be the first to share your thoughts.